Professional Documents
Culture Documents
FERNAND GOBET
First published 2019
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2019 Fernand Gobet
The right of Fernand Gobet to be identified as author of this work has
been asserted by him in accordance with sections 77 and 78 of the
Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or
registered trademarks, and are used only for identification and
explanation without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Names: Gobet, Fernand, author.
Title: The psychology of chess / Fernand Gobet.
Description: New York : Routledge, 2019. | Series: The Psychology
of Everything
Identifiers: LCCN 2018020887 (print) | LCCN 2018024430 (ebook) |
ISBN 9781315441863 (ePub) | ISBN 9781315441870 (Adobe) |
ISBN 9781315441856 (Mobipocket) | ISBN 9781138216631
(Hardback) | ISBN 9781138216655 (Paperback) |
ISBN 9781315441887 (eBook)
Subjects: LCSH: Chess—Psychological aspects.
Classification: LCC GV1448 (ebook) | LCC GV1448 .G634 2019 (print) |
DDC 794.1019—dc23
LC record available at https://lccn.loc.gov/2018020887
ISBN: 978-1-138-21663-1 (hbk)
ISBN: 978-1-138-21665-5 (pbk)
ISBN: 978-1-315-44188-7 (ebk)
Typeset in Joanna
by Apex CoVantage, LLC
To the Queens of my Life
Chess is life.
World champion Bobby Fischer
Preface xi
Acknowledgements xiii
Opening 1
2 Chunks! 13
11 Endgame 117
of the game is not necessary to enjoy this book, and care has been
taken not to use chess jargon, or to explain it when it is used. No
background in psychology is expected, either.
In line with this collection, the number of references is kept to
a minimum. The suggested items in the Further Reading section at
the end of the book provide sources where detailed pointers to the
literature can be found.
Around the time of my PhD, I had the very good fortune of col-
laborating with Adriaan de Groot and Herbert Simon, two outstand-
ing scientists who made seminal contributions to the psychology
of chess. Later, I trained several gifted researchers (Merim Bilalić,
Guillermo Campitelli, Philippe Chassy and Giovanni Sala) who had
decided to devote their PhD research to aspects of chess psychology.
Both the old and new research will be described in this book.
In my first career, I was a professional chess player, reaching the
title of an international master and playing for the Swiss national
team. Therefore, I have sometimes taken the liberty of providing first-
hand evidence and insights about chess players and their mind.
ACKNOWLEDGEMENTS
Chess fascinates many people, even those who do not play it. It is
played at all ages, in all countries, without any attention to religion
or ethnic background. As a symbol of intelligence and good decision-
making, it is often used in advertisements for banking and business.
As a symbol of genius bordering on madness, it is a common theme
in fiction and has generated numerous books and movies.
Born in India in the 6th century, chess was adopted in Persia (Iran)
in the next century, and then by the Arabs. The first books on chess
were written in the 9th century, but did not survive to our day. It is
also in the 9th century that chess was introduced to Europe, through
Spain and Sicily. By the 13th century, chess was the dominant game
in Europe, as witnessed in the Book of Games, written in 1283 at the
request of Alfonso X of Castile: “Since chess is the noblest game,
which requires the most skill compared to all the other games, we
are going to talk about it first of all”. By then, its influence had spread
well beyond the sphere of games and it was part of culture at large.
For example, in the second half of the 13th century, Jacobus de Ces-
solis, an Italian monk, preached morality using chess as metaphor:
“In life, as on the chessboard, each piece has its own rights but also
its own obligations”.
2 OPENING
discussion of key positions and the way to handle them for the theory
of endgames. There were some attempts to identify the fundamental
principles of play, most notably by Wilhelm Steinitz, the founder of
the classical school of chess, with his 1889 textbook The Modern Chess
Instructor, and Aron Nimzowitsch, one of the founders of the hyper-
modern school, with his 1925 book My System. However, despite their
great originality, both fell short of the rigour of a scientific theory. In
addition, what is lacking generally with chess is any effort to test its
“theories” in a systematic way, beyond recording new games. Such
tests are obviously at the heart of scientific research.
Chess has been the topic of much scientific research. It has been
investigated by a number of academic disciplines, including sociol-
ogy, ethnology, philosophy, mathematics and neuroscience. By far, it is
in computer science (including artificial intelligence) and psychology
that chess has been studied to the greatest effect. In artificial intelli-
gence, chess has been a standard task for the development of machine
learning and search algorithms. In psychology, it has been the topic
of seminal research into perception, memory, learning, thinking and
decision-making. It has sometimes been called the drosophila of cogni-
tive psychology, by analogy to the role of the fruit fly in genetics.
logical principles. They also provided a list of the physical and mental
qualities required by chess, which were instrumental in convincing
the Soviet government that chess should be encouraged as an activity
leading to the development of self-discipline and the improvement
of intellectual competences.
However, it is to Dutch psychologist and chess master Adriaan de
Groot that we owe the first experimental study on chess psychol-
ogy, which he carried out for his PhD research. In 1938, he was
earning money as a journalist by covering the AVRO tournament in
Amsterdam, which brought together the world’s best eight players. He
managed to convince five of the participants to take part in his experi-
ments, including world champions Alexander Alekhine and Max Euwe.
Amusingly, some of the data were collected after the tournament on the
steamer carrying many European masters to Buenos Aires, where the
1939 Olympiads (world championship by teams) were held. As the trip
was rather long, players were grateful to participate in these experi-
ments and therefore to break the monotony of the voyage.
De Groot studied not only chess players’ ability to find good
moves, but also their ability to rapidly understand the gist of a posi-
tion even after seeing it just for a few seconds, as well as their ability
to memorise these positions rapidly and accurately. Many of the ideas
presented in this book can be traced backed to de Groot’s PhD thesis.
The second key study in chess psychology, carried out by Herbert
Simon and William Chase in 1973 at Carnegie Mellon University in
Pittsburgh, developed a powerful theory – called chunking theory –
to explain de Groot’s data. Combining experimental methods with
ideas from artificial intelligence and computational modelling, Simon
and Chase performed a series of experiments that inspired much of
the research carried out in the following decades.
Nowadays, chess psychology is an active domain of research and
is arguably still the main domain in expertise research. Many dif-
ferent aspects of chess are studied, from cognition to personality to
intelligence. Several reasons explain this popularity, including: chess
has its own rating system, the Elo rating, which offers a precise and
up-to-date measure of skill; it has an ideal balance between simplicity
OPENING 5
PREVIEW OF BOOK
The following three chapters cover standard topics in cognitive psy-
chology: perception, learning, memory and decision-making. They
are anchored in the work of de Groot, Simon and Chase. Chapter 4
discusses the relative roles of practice and talent, and emphasises the
importance of studying their interaction. Chapter 5 tries to under-
stand why men appear to perform better than women in chess at the
top level. The following three chapters are devoted to applied chess
psychology, and cover topics such as errors, style, intuition, train-
ing, psychological warfare and cheating. The discussion of style and
intuition will draw on recent developments in artificial intelligence,
focusing on what they tell us about human psychology. Chapter 9
examines the possible benefits of playing chess (e.g. for education or
psychotherapy) and Chapter 10 addresses the potential costs (e.g. the
hypothesis that there is a link between madness and chess).
Anybody who has seen chess masters playing bullet chess (1 minute per
side for the entire game) or simultaneous exhibitions, where they play
against 30 or 40 opponents at the same time, would have been struck by
their amazing ability to play good moves very quickly. Indeed, the qual-
ity of moves played under these taxing conditions is surprisingly high,
although not as high as with games played under normal conditions (on
average, 3 minutes per move). It is as if masters see the board differently
than weaker players. Where novices see wooden or plastic pieces, masters
see trajectories, ideas, concepts and sequences of moves. In fact, the same
applies in other fields: one of the hallmarks of experts in science, medicine
and sport is the ability to rapidly perceive the key features of a problem.
The first person to have addressed this question empirically was
Adriaan de Groot in his doctoral dissertation, originally published
in Dutch in 1946 and translated in English in 1965. Because of the
number of issues it addressed and its strong scientific impact, this
work has become a classic in psychology.
positions when they were looking ahead and that they anticipated
longer sequences of moves – that is, they were searching deeper. In a
first experiment, he gave chess players a board position unknown to
them and asked them to select what they thought was the best move.
He also asked them to say aloud what they were thinking about.
The players consisted of amateurs, candidate masters and world-class
grandmasters, including world champions. The transcripts of play-
ers’ utterances – called verbal protocols – were then analysed in great
detail, both qualitatively and quantitatively.
The results did not support his expectations: although grandmas-
ters played better moves, they did not differ substantially from other
players with respect to structural variables such as the depth of search,
the number of moves considered or the strategies used when carrying
out search (see Chapter 3 for details). There was an important dif-
ference, however. The best players were able to pinpoint promising
solutions very rapidly, which allowed them to narrow down their
search drastically. As de Groot put it, the world champion understood
the problem position better after 5 seconds than a candidate master after
15 minutes! This was fully unexpected. What was critical was not the
detail of the way players analysed the position by trying out different
moves, sometimes for more than 30 minutes. Rather, the difference
resided in the very first few seconds of seeing a position: perception
is central in chess expertise.
In a second experiment, de Groot directly tested this hypoth-
esis. He presented a position briefly, from 2 to 15 seconds, took it
away from participants’ view, and asked them to reconstruct it as
precisely as they could. As expected, grandmasters did much bet-
ter than candidate masters, who in turn did better than amateurs.
Whilst a grandmaster could reconstruct nearly the entire posi-
tion correctly, a strong amateur struggled to remember half of the
pieces.
De Groot explored several variants of this experiment. In some
versions, he asked players to think aloud, either during the presenta-
tion of the position, immediately after or 30 seconds after. From the
protocols, it is clear that experts did not see individual pieces, but
THE EYE OF THE MASTER 9
the position, with the details omitted. In the 1960s, there was an
interesting debate about this issue between Soviet psychologist Oleg
Tikhomirov and Herbert Simon, with Tikhomirov defending the view
that perception is holistic, while Simon argued that local mecha-
nisms (e.g. perception of relations of defence and attack between
pieces) were sufficient for explaining the data. More recently, Gobet
and Chassy run computer simulations based on the idea of chunks
and templates (see the next chapter), showing that experts’ percep-
tion, even though it might look holistic, can be accounted for by the
incremental construction of an internal representation using patterns
that are initially fairly small.
Another idea proposed by de Groot seems better supported by the
empirical evidence. He suggested that chess masters used anticipatory
schemas. These dynamic schemas contain information allowing play-
ers to anticipate potential actions. As experts have more and better
developed schemas, they can anticipate actions better. Vincent Fer-
rari and colleagues at the University of Provence (France) tested this
hypothesis. In a first experiment, players saw two positions in quick
succession and had to say whether the second position was the same
as the first one. The results showed that strong amateurs performed
better when the two positions appeared as a normal sequence of
moves, unlike beginners who could not use information about the
normality of moves.
In a second experiment, Ferrari and colleagues studied whether
players tend to recall positions as they were shown or, rather, the posi-
tions that would occur after the standard move is played, as predicted
by the presence of anticipatory schemas. A recognition task was used.
Players saw 10 chess positions displayed in succession; half of the
positions were standard opening situations, while the other half were
a different set of opening situations, this time with one additional
move played. During the recognition phase, players were presented
with 10 old positions (the positions they had seen in the first phase
of the experiment) and 10 new positions (half were the positions
they had seen plus one standard move, and the other half were the
positions they had seen minus the standard move). The results showed
12 T H E E Y E O F T H E M A S T E R
that the strongest players (class A players) made many false recogni-
tions, where they recalled not the position they had seen, but the
position after the normal move had been made. The beauty of this
experiment is that better players committed more false recognitions
than weaker players, showing that in some circumstances expert per-
ception can lead to errors. In sum, these two experiments back up the
hypothesis that experts use anticipatory schemas in their perception:
rather than recalling a scene the way they saw it, experts tend to recall
it the way it normally unfolds in the near future.
PERCEPTION IS COGNITION
The importance and speed of perception is not limited to chess, but
has been documented in many other domains of expertise, such as
music, medicine, sports and driving. In all these domains, experts lit-
erally see a different problem situation and categorise it in a better way.
Rather than being innate, experts’ perception is the product of many
years of practice and study. One of de Groot’s major contributions is
to have shown that there is no clear boundary between perception
and cognition: in chess and in other domains, perception, memory
and problem-solving are tightly interconnected.
2
CHUNKS!
Figure 2.1 Chunks learnt by a weak player (left diagram) and a master (right
diagram).
CHUNKS! 15
RECALL EXPERIMENT
Simon and Chase’s genius was not only to have proposed a power-
ful theory, but also to have supported it empirically with elegant
experiments. They focused on de Groot’s perception task – which
they considered as a recall task – providing both a replication and an
extension. A chessboard was shown for five seconds, after which play-
ers attempted to reconstruct it. There were two important differences
in comparison to de Groot’s experiment. First, Simon and Chase did
not ask participants to think aloud during the experiment. Second,
in addition to board positions taken from masters’ games, they used
random positions, where the pieces of a game position were haphaz-
ardly placed on the board. The results showed that there was a skill
effect with positions taken from games, but not random positions.
CHUNKS! 17
COPY TASK
In comparison to the original study, another important addition was
a copy task, which provided critical information about the structure of
chunks. In this task, players could still see the stimulus board when
the position was reconstructed on a second board. So, typically, play-
ers would glance at the stimulus board, direct their attention to the
reconstruction board and place a few pieces, glance again at the stim-
ulus board, and place yet more pieces. Simon and Chase used the fact
that the two boards could not be fixated at the same time to provide
an empirical operationalisation of a chunk: pieces replaced together
after a glance at the stimulus board belong to the same chunk. Thus,
complementing the first definition of a chunk as a group of pieces,
we now have here a second definition.
Simon and Chase also studied the latencies between the placements
of two pieces. They reasoned that, if two pieces belong to the same
chunk, they should be replaced rapidly together. By contrast, if two
pieces belong to two different chunks, the latency between them
should be longer. As predicted, most of the pieces within a chunk – as
defined by whether players glanced or not at the stimulus board –
were replaced in less than 2 seconds, while the pieces that belonged
18 C H U N K S !
condition, the pieces were dictated file by file; in the final condition,
the pieces were presented randomly. In line with chunking theory,
players recalled the positions best when the pieces were dictated using
chunks. Random order yielded the worst recall. In another experi-
ment, Charness presented the pieces visually, with the same results.
These experiments are important, because the variable of interest –
whether pieces belonged to the same chunk or not – was controlled
by the experimenter.
also interferes with the memory of the positions that come after (pro-
active interference). Again, chunking theory predicts that recall of all
positions should become increasingly poorer as the number of boards
increases. Several authors carried out the experiment, and the results
showed that, while the task was too hard for amateurs, masters can
do it relatively well, even though there seems to be a barrier after
five boards.
Figure 2.2 Examples of a position taken from a game (left diagram) and a random
position (right diagram).
a game position, and randomly placing them on the board (see Fig-
ure 2.2 as an example). Remember also that, with random positions,
Simon and Chase did not find any difference in performance between
their master, class A player and novice. This pattern of results – vast
skill differences with game positions, but no differences with random
positions – was both spectacular and intuitively satisfying: experts can
use their knowledge with structured material but not with unstruc-
tured material. It thus found its place in numerous textbooks and
popular-science books. The problem is that the second part of the
argument (no skill effect with random positions) is not true.
When I was developing CHREST in collaboration with Simon,
the programme consistently and stubbornly predicted that masters
should remember random positions better than weaker players, even
though the absolute number of pieces remembered should be much
smaller than with game positions. In one of the rare cases where his
scientific intuition let him down, Simon did not believe the simula-
tions. He thought there were mistakes in the computer programme.
There were small mistakes, for sure, but once corrected, the behav-
iour of the programme was the same: it predicted a skill effect with
random positions. The mechanism behind this behaviour was actu-
ally simple and a direct consequence of chunking theory. By chance,
there are patterns even in random positions, albeit not many of them.
CHUNKS! 25
Masters, who have acquired more chunks than weaker players, are
more likely to recognise one of these patterns by chance and thus
access a chunk in long-term memory. Thus, they should show a small
superiority over weaker players.
We combed the literature to find all studies that had used random
positions in experiments on chess memory. These positions were typi-
cally used as a control task to ascertain that better players do not have a
better memory in general. When put together, the results indicated that
there is a skill effect even with random positions, even though it was
smaller than with game positions. On average, masters recalled about
20 pieces with game positions and five pieces with random positions,
while weak amateur (below class B players) recalled five pieces and
2.5 pieces, respectively. In fact, a skill effect was found in 12 out of 13
studies, the exception being Simon and Chase’s study! The skill effect
did not reach statistical significance in most studies, due to their small
sample sizes. In the end, Simon conceded that he should have believed
the computer simulations rather than his own intuition.
This result is important theoretically. Not only is it a direct predic-
tion of chunking theory, but it is actually difficult to explain for most
theories of expertise – for example, theories that assume that experts’
knowledge is primarily coded by relatively high-level memory struc-
tures such as schemas. The reason is that small memory structures
such as chunks seem necessary for explaining this result.
There is very recent twist in the story. In a meta-analysis of the
studies having used random material not only with chess, but with
other domains of expertise as well, my PhD student Giovanni Sala
found that this skill effect generalises to most domains of exper-
tise, thus providing additional support for the existence of chunking
mechanisms in human cognition.
BLINDFOLD CHESS
There is a variant of chess that supremely taxes memory: blindfold
chess. With blindfold chess, one plays without seeing the board,
unlike the opponent who, typically, can see it. Players normally com-
municate using the algebraic notation (e.g. 1. d4 Nf6 2. c4 g6, etc.),
which is the standard way of recording moves. As if playing one
blindfold game was not difficult enough, it is also possible to play
several blindfold games at the same time. Whilst most players above
candidate master level can play one blindfold game without too
CHUNKS! 27
Take Attack
Retreat White
rook queen
rook
PROGRESSIVE DEEPENING
The qualitative analyses describe, in great detail, the way the deci-
sion processes are structured. Most of de Groot’s book is devoted to
them. An important outcome was that all players were investigating
the same base move (one of the first moves in the problem situation)
several times – a process that de Groot called progressive deepening. The
reinvestigation could be immediate, or separated by the analysis of
another base move. In either case, reinvestigating a base move makes
it possible for the player to study it with more precision, by increas-
ing the number of moves anticipated or by improving the evaluation
of the terminal position.
At first blush, this process seems repetitive and inefficient, since
one might think that resources could be better used by analysing
different base moves and new sequences of moves. However, progres-
sive deepening is actually quite adaptive. First, it helps alleviate the
restrictions imposed by the limited capacity of short-term memory.
Revisiting the same sequence of moves several times increases the
likelihood that it is stored in long-term memory, thus making analysis
THE BEST MOVE 33
nine matches he had played at the time, Kasparov won eight of them;
he lost only his first match, presumably because he was new to this
type of event. His median performance was 2646 Elo, which was
the strength of a top grandmaster at the time and would still have
placed him in the best six players in the world. In fact, Kasparov’s
performance in these matches was typically less than 100 points
below the level displayed in normal tournament play. For example,
his rating in July 1987 was 2735.
The teams consisted of four to eight strong masters and grandmas-
ters, and thus Kasparov’s thinking time was reduced proportionally. As
search is carried out serially and relatively slowly (perhaps 10 posi-
tions per minute, based on de Groot’s data), reducing thinking time
should affect the number of positions examined and consequently
decrease performance. The loss should not be as large if pattern rec-
ognition played a larger role, because this process is assumed to occur
rapidly. That is what was observed in the actual results. In line with
this analysis, there was no correlation between Kasparov’s perfor-
mance and the number of opponents he was facing, which directly
affected the time available for his thinking.
While I still believe that this analysis is correct, two factors should
be taken into account. The first is that Kasparov did carry out a fair
amount of search, as is obvious from looking at the games he played
in these matches. The opposition essentially consisted of professional
players, including many grandmasters; therefore, he could not just
play normal moves and wait for the opponent to make a mistake. He
had to provoke such mistakes by complicating games, which he did
with brio. The second factor, related to the first, is that Kasparov was
extremely well prepared for these encounters, with the exception of
the first one, as noted above. He would analyse at least 100 games
of each of his opponents to identify their strengths and weaknesses.
Then, he would use this knowledge to steer the play into the kind
of positions that a specific opponent did not handle well and thus
increase the chances for that opponent to play inferior moves. For
example, if an opponent was relatively weak in anticipating a sud-
den attack in what seems a quiet strategic position, Kasparov would
THE BEST MOVE 41
and 27 days – that is, only five years and four months of deliber-
ate practice! Carlsen is actually not known to be the most assiduous
worker – for example, he prefers watching or playing football to
studying chess. In a 2014 study with my PhD student Morgan Ereku,
I estimated the amount of time that the top 11 players in the world
had engaged in deliberate practice. We found that Carlsen’s number
of years of practice was significantly less than the average of the other
players. In spite of this, the difference between Carlsen and the second
player in the world (Levon Aronian) was 66 Elo points, which was
about the same as the difference between Aronian and the 14th player
in the world (Anish Giri)!
TALENT IN CHESS
As noted at the beginning of this chapter, the main alternative expla-
nation to practice is talent. Theoretically, there are good reasons to
believe that there are innate differences in cognitive abilities in gen-
eral, and in those underpinning chess skill in particular. From an
evolutionary point of view, variability is necessary for evolving not
50 P R A C T I C E M A K E S A L M O S T P E R F E C T
Environment
Practice Intelligence
Performance
Environment
Practice Intelligence
Performance
Practice Intelligence
Figure 4.1 The spaghetti model, which emphasises interactions between environmental and innate factors.
54 P R A C T I C E M A K E S A L M O S T P E R F E C T
who got the title at 15 years, 6 months and 1 day. At one point ranked
8th in the world, she has won games (in either classic or rapid chess)
against many of the top players of the world, including Anatoly Kar-
pov, Garry Kasparov and Magnus Carlsen.
Sofia did not reach as high a level as her sisters. In an interview,
Susan noted that Sofia was the most talented of the three, but also
the laziest. That being said, when Sofia was 14 years old, she won a
tournament in Rome ahead of several leading grandmasters, with the
stunning score of 8½ out of 9 and a performance of 2879 Elo, which
was at the time one of the highest in chess history.
The Polgár sisters also obtained remarkable results when playing
together for Hungary, which might be called Polgáry, as they made
up 75% of the four-player team. They won the women’s chess Olym-
piads in 1988 and 1990, breaking a long series of victories by the
Soviet Union.
László Polgár’s experiment yielded clear results: his method of
education, which emphasises specialising in a field from an early
age, produced top-level chess players. But does it really prove that
nurture is more important than nature? No, because the design of the
experiment lacks a key feature: random selection of the participants.
Thus, genetic influences are possible, even likely: László and Klara
each have a PhD degree in education. A better design would be to
raise randomly selected children. In fact, Dutch billionaire Joop van
Oosterom proposed to do just this: he suggested to László and Klara
Polgár that they could use their training methods with children from
a developing country. László was interested, but Klara declined, prob-
ably wisely. As Susan Polgar put it in a Psychology Today interview: “[My
mother] understood that life is not only about chess, and that all the
rest would fall on her lap”.
EXPLANATIONS
Gender differences in chess performance are not disputable. But what
are the reasons? Are they the same as those proposed to explain gen-
der differences in science, technology, engineering and mathematics
58 M E N V S . W O M E N
Biological explanations
Socio-cultural explanations
the opponent was a man; and (c) low gender stereotype, where par-
ticipants were told that the opponent was a woman. In the last two
conditions, the gender stereotype was made explicit by telling players
that men play better chess than women. Women performed worse than
men in the high stereotype condition, but not in the other two con-
ditions. However, the hypothesis of stereotype threat was not sup-
ported in another study. In an analysis of over 5.5 million competitive
games in international tournaments, Tom Stafford did not find any
support for the stereotype threat hypothesis. Contrary to expectation,
women actually performed better against men than expected from
their ratings.
Psychoanalytical explanations
WHAT IS INTUITION?
Intuition can be defined as the ability to understand a situation rapidly
and effortlessly. It is a topic that has been of great interest both to
psychologists and chess players. The former consider it as one of the
defining features of expertise – think of a medical doctor able to diag-
nose a disease almost immediately. The latter wonder how a player
like Magnus Carlsen is so good at finding good moves, often with-
out much calculation, while weaker grandmasters cannot find them
despite considerable thinking. Also, how is it that the same concept –
intuition – is used to describe players with styles as opposite as world
STYLE AND INTUITION 71
taken as evidence that experts can still display fluid behaviour despite
lacking time for planning. This is a reasonable conclusion, although it
hard to evaluate the experiment given the few details provided. Unfor-
tunately for the theory, when a proper experiment was carried out by
Robbins and colleagues, where players had to generate random num-
bers while solving tactical chess problems, it was found that players
were affected by the interfering task. With the group of stronger play-
ers, performance dropped by about one third when compared to the
control condition where they simply solved chess problems.
Computer intuition
that computers not only play chess better than humans, but that they
understand it more deeply and have a much better intuition. This
is a repulsive thought for most human players, but it is probably a
reality now. Most computer programmes playing chess are based on
powerful evaluation functions, combining dozens of features, such as
safety of the king, mobility of the White-square bishop and control
of centre, with the ability of searching billions of positions before
selecting a move. Still, it is totally out of the question to calculate all
possibilities, and thus computers rely on some form of “intuition” to
make a move. In fact, there are many examples of surprising moves
played by computers, which turned out after extensive human analy-
ses to be very deep positionally. In 2006, world champion Vladimir
Kramnik lost 4–2 against Deep Fritz, which actually was not even the
best computer programme at the time. In the 6th game, Deep Fritz
played a rook manoeuvre that pundits derided as child-like. The next
moves of the game showed that this manoeuvre was the prelude
to a very deep plan that led to fatal weaknesses in Kramnik’s king’s
side and ultimately to material loss.
If there were any doubts that artificial intuition is possible, these
were recently shattered by AlphaGo’s victories in Go. Just before the
entrance of AlphaGo into the Go scene in 2015, artificial intelligence had
progressed very slowly with this game, to the point that some
experts thought that computers would never be able to beat the best
humans. The most optimistic researchers were of the view that it
would take a computer at least 10 years to win a game against a
Go professional player. Thus, before the match against South Korean
grandmaster Lee Sedol, spirits were high in the Go community and it
was expected that AlphaGo would be beaten easily. After all, Lee Sedol
was only one of two players to have won 18 international titles (more
or less the equivalent to Grand Slam titles in tennis). Unfortunately
for humankind, Lee was thrashed 4–1.
AlphaGo primarily used three artificial intelligence techniques:
deep learning, reinforcement learning and Monte Carlo tree search.
Deep learning consists of sophisticated methods for adjusting the
weights of an artificial neural network, using grandmaster games as
STYLE AND INTUITION 75
complex moves. These criticisms might be true, but the fact is that we
have now a super-program that learns by itself from scratch and that
plays much better than the best humans. In addition, AlphaZero has a
very intriguing and intuitive style, sacrificing material for long-term
positional compensations that most humans cannot see. Just like Go,
AlphaZero seems to show that human understanding of chess, in spite
of centuries of practice and study, is rather limited.
AUTOMATISMS
Automatisms are important in chess, since they allow one to make
decisions rapidly. Some consist of very simple conditioned actions,
such as “if your opponent takes one of your pieces, take it back imme-
diately”, or “if a piece is attacked, move it away”. Others are subtler
and directly link to the chunking mechanisms discussed in chapters 2
and 3. For example, given a certain pawn structure, it is often a good
idea to place one’s knight in front of an isolated pawn. The beauty
of chess is that it is a game of exceptions, and even the best proce-
dural knowledge will be incorrect every so often. If one follows one’s
ERRARE HUMANUM EST 79
CHESS IMAGES
Krogius spends a great deal of time with what he calls “chess images”.
The term is maybe not the best one, as these images contain not only
visual information on the location of pieces, but also more concep-
tual information such as the evaluation of the position. In some of
Krogius’s examples, images really refer to what I have called sche-
mas or templates in previous chapters. Chess images are very use-
ful in most cases, as they provide much information. They can also
produce powerful negative effects, however. For example, facing an
unexpected move by the opponent, a player will often still think
using the mind-set provided by a chess image that was suitable just
one move ago, and might miss a new opportunity. Several examples
will be provided below.
Another class of errors, which Krogius calls “retained images”,
concerns the case where, when anticipating a sequence of moves, a
piece taken by the opponent somehow still remains present in the
mind’s eye. The difficulty thus resides in correctly updating the loca-
tion of pieces or their disappearance in the mind’s eye. For example,
with this type of error, a player would defend in his calculation against
a bishop that actually is not on the board anymore. In a related type
of errors, which Krogius calls “inert images”, the player automati-
cally carries a positive evaluation reached at some point of the game
on to the following moves and is essentially on an automatic pilot
mode. As a consequence, the difficulty of winning the game can be
underestimated and one’s sense of danger blunted. So, for example
in the game Petrosian against Korchnoi in the 1963 Moscow tourna-
ment, Petrosian had had a winning position since the opening. Still
winning after 34 moves, he overlooked a simple combination and
played a careless move, which lost immediately. This type of error is
related to overconfidence, which I shall discuss towards the end of
this chapter. It is also related to the Einstellung effect that I described
80 E R R A R E H U M A N U M E S T
EMOTIONAL FACTORS
It might come as a surprise for non-players, but chess can be a very
emotional game. To begin with, most players have their bêtes noires,
opponents against whom their score is much less than expected by
their relative objective strength. For example, Soviet grandmaster
Efim Geller was the bête noire of Bobby Fischer, with five wins, three
defeats and two draws, in spite of Fischer’s superiority. Similarly,
American grandmaster Hikaru Nakamura, currently number 8 in the
world, struggles against Magnus Carlsen, having lost 12 games and
won only one, with 21 draws. Krogius reports that he analysed 80
games between 10 pairs of players where one player was a bête noire
for the other. He found that the players on the losing side commit-
ted more obvious strategic errors and tactical mistakes against their
bête noire than against other players. Krogius argues that a negative
emotional state strongly diminishes these players’ vigilance and ner-
vous resistance, although it could be argued that, in addition to this,
there might be a profound incompatibility in style. For example, in
the case of Fischer and Geller, Geller was known for his very danger-
ous, uncompromising attacking style. Interestingly, bêtes noires are
not transitive. For example, ignoring draws, Tal had a lifetime score
of 8–5 against Bronstein, and Bronstein had a lifetime score of 7–4
ERRARE HUMANUM EST 81
INSUFFICIENT KNOWLEDGE
Many errors are caused by a lack of knowledge. At weaker levels, this
can be due to not knowing typical openings, standard tactical and
strategic patterns in the middle game and common manoeuvres in
endgames. As noted in an earlier chapter, masters and grandmasters
often win games by waiting for mistakes, and lack of knowledge is
at the core of many of them. Mistakes due to insufficient knowledge
also occasionally occur at the top level. For example, in the first game
of the world championship held in 2010 in Sofia between Veselin
Topalov and Viswanathan Anand, world champion Anand could not
remember the move order of a variation he had prepared in depth. He
chose the wrong order and got thrashed by his opponent, who had
remembered the brilliant combination that refuted Anand’s move.
TIME TROUBLE
Competitive games are played with a clock, and thinking time is lim-
ited. The limit is strict and overstepping one’s thinking time means
defeat by “losing on time”, which is not uncommon. An extreme
case is grandmaster Friedrich Sämisch, one of the leading players in
the 1920s, who lost every single game on time in two tournaments
in 1969, when he was 73 years old. The exact time limits vary from
tournament to tournament, but there has been a tendency towards a
decrease of thinking time over the years, partly to reduce the duration
of games and make them more exciting. Currently, many tournaments
use time controls of 90 minutes for the first 40 moves, followed by
ERRARE HUMANUM EST 83
I chose a very sharp variation and had the advantage after the opening.
For several moves, I considered a pawn sacrifice in the centre in order
to initiate a direct attack against Spassky’s king. However, the idea
never worked satisfactorily, so I progressively directed my attention
to another central break. Ironically, at the very moment where I gave
up on my initial idea, it would have in fact been the winning move.
room. Then, one should look again at the board, with fresh eyes, so
to speak. This should make it possible to look at the position as it is
on the board right now. The advice is now to stare at the position for
about a minute, through the eyes of a novice, and ask very basic ques-
tions such as: is my queen attacked? Is there a direct threat against my
king? According to Kotov, this double-checking procedure drastically
decreases the risk of committing a blunder.
This is very good advice. Having used it in my chess career, I can
vouchsafe that it saved me from an embarrassing oversight more than
once. Unfortunately, one important part of Blumenfeld’s rule is not
possible anymore. The International Chess Federation has changed its
rules, and moves must be written down on the score sheet only after
they have been executed on the board.
8
PSYCHOLOGICAL WARFARE
AND TRAINING TECHNIQUES
simplify positions by trading pieces off and then rely on his technique
to defend them. This approach was new to Alekhine, who in the past
had used more risky and active methods of defence. However, the
psychological surprise turned out to be more important than Alekh-
ine’s relative inexperience with this type of position.
THEORETICAL NOVELTIES
Knowing the “theory of openings” is essential for chess players. It is
not a “theory” in the scientific meaning of the term – a set of mecha-
nisms explaining a body of empirical phenomena. It is not even a set
of principles. Rather, it is the compilation of moves that have been
played in the past and their evaluation: did White or Black have the
advantage, or was the position equal? This body of knowledge has
grown rapidly in the last decades, and it is not uncommon for play-
ers to blitz 20 moves or more in the opening, playing them from
memory. In a paper with Philippe Chassy, we estimated that masters
must have memorised about 100,000 opening moves.
Then comes the much-anticipated moment where the game diverges
from theory: a theoretical novelty. At the amateur level, novelties often
reflect players’ lack of knowledge. By contrast, at the professional level,
many theoretical novelties are the fruit of home preparation, sometimes
weeks or months before the actual game. In the quietness of their living
room, players can spend hours and hours studying a position, moving
the pieces on the board and even asking the advice of their colleagues.
In the last two decades, players have increasingly used computers to
help their investigations, which has multiplied the number of positions
that can be studied by several orders of magnitude. As all this is done
before the start of the game itself, it is legal. The aim of these analyses
is not only to find the move that is objectively the best, but also to find
moves that will confront the opponent with the kinds of positions
they do not like or that are unknown to them. This can occasionally
include inferior positions, especially with Black, with the idea that the
opponent will not be able to find the winning plan in the limited time
allowed by competitive games.
92 P S Y C H O L O G I C A L WA R F A R E A N D T R A I N I N G
OUTRIGHT CHEATING
The last category consists of methods that clearly constitute cheat-
ing. Bobby Fischer accused Soviet grandmasters of such practices:
according to him, they were rapidly drawing amongst themselves
to conserve energy, were consulting during games and were even
fixing games amongst themselves in tournaments. A 2006 study by
Moul and Nye supports Fischer’s view. Using statistical analyses, they
showed that the rate of draws amongst Soviet grandmasters was higher
in important international tournaments than in national tournaments.
Collusion is not uncommon nowadays. The classic example is the
last round of a tournament, where a win would lead to prize money
for one of two players facing each other, but a draw would not. The
temptation is great for the two players to come to an agreement before
the game, with, for example, the player losing the game getting more
money than the winner. On a slightly more bizarre note, a number
of tournaments were simply invented, for example in Rumania and
Ukraine, with no games actually played, so that some players could
win Elo points or be awarded the title of international master or even
grandmaster.
Technology has created new opportunities for cheaters. Several
players have been caught, during a competitive game, using a chess
programme on their smartphone during breaks in the bathroom.
One player, called the “James Bond of chess”, hid a transmitter in
his shoe in order to receive move suggestions sent by a friend. In the
2010 Olympiads, three French players used a sophisticated scheme
for communicating moves. The first player took advantage of live
Internet broadcasting of the tournament to follow the game from his
home and analyse the current position with a computer programme.
He would then text the best move to the second player, who would
communicate it to the player actually playing the game, using his
standing or sitting at various places as a code. The scheme was subtle,
but had a not-so-subtle flaw: they used a mobile phone of the French
Federation, which handed down heavy suspensions to the three play-
ers. To make such technological cheating harder, the International
94 P S Y C H O L O G I C A L WA R F A R E A N D T R A I N I N G
Chess Federation has banned the use of mobile phones and other
electronic devices in playing venues. In the same spirit, tournament
organisers now increasingly add a lag to the transmission of moves
on the Internet.
factors rather than the way search is carried out, such as an inability
to make rapid decisions.
Kotov’s book also comes with various exercises for improving
one’s ability to visualise a position and increase depth of search. Most
of these exercises boil down to finding the best move in a given
position, and are certainly useful. Other authors have proposed to
play games blindfold, again to improve one’s visualisation skills. In
the paper with Peter Jansen mentioned earlier, we were rather criti-
cal about this piece of advice. Our argument was that finding good
moves and being able to calculate deep variations was primarily the
consequence of having acquired considerable knowledge, which
allowed players to identify plausible moves rapidly through pattern
recognition, not only in the problem situation but also in the posi-
tions visualised during look-ahead search. Another point was that, if
one has a good understanding of a position, it is often not necessary
to anticipate moves very deeply. However, the paper was written for
players aiming to become masters, and one could argue that, when
one wants to reach higher levels of skill such as grandmaster, playing
blindfold games might help fine-tune one’s calculating skills.
9
THE MAGIC BULLET?
While most players agree that chess is a fascinating game with endless
possibilities and graced with real aesthetic value, there is disagree-
ment about whether it ultimately benefits or penalises those who play
it. Siegbert Tarrasch famously wrote that “chess, like love, like music,
has the power to make men happy”. Some have highlighted possible
benefits for education. A quick look at the Internet will display many
slogans such as “Chess makes kids smarter” and “Chess improves
mathematics”. Others have directed attention to the potential costs
of playing chess. Orson Wells argued that to destroy a person, you
should teach her chess, and George Bernard Shaw asserted that chess
“is a foolish expedient for making idle people believe they are doing
something very clever when they are only wasting their time”. Rare
are the activities that have simultaneously inspired such extreme views
about their merits and demerits. This chapter deals with the presumed
benefits, and the following chapter with the presumed costs.
Numerous websites repeat these claims and some authors use them as
arguments for advertising the practice of chess. In 2001, the US Chess
Trust asked me to review the available evidence, objectively and with-
out any biases. I carried out this review with Guillermo Campitelli.
This is obviously a very important topic for the chess world, and a
positive answer would have significant implications for education. In
fact, it would be a very elegant solution to the difficulties that children
and teenagers currently encounter when studying STEM topics.
This question more generally relates to the issue of transfer: can
abilities acquired in a particular field (e.g. geometry) generalise to
other fields (e.g. science, music) or general abilities (e.g. intelligence,
reasoning)? Broadly speaking, there are two types of transfer. With near
transfer, the two fields overlap, such as geometry and algebra. With far
transfer, there is no or little overlap between the two fields. An example
would be geometry and English. With chess, the strong claim is that
there is far transfer.
Establishing the presence of far transfer and thus the causal role of
chess raises interesting methodological questions. Three groups are
necessary: a treatment group (playing chess, in our case), an active
control group (e.g. playing checkers) and a passive control group (no
treatment). The active control group is needed in order to control
for unspecific factors such as placebo effects and expectations. In
addition, participants should be allocated randomly to the three
groups. This avoids systematic biases, such as the case where all the
more intelligent children are allocated to the chess group whilst the less
intelligent children end up in the other two groups. Measures should
be taken before the experiment to ensure that the groups do not dif-
fer at the outset, and of course after the experiment, to evaluate the
effect of treatment. To measure potential changes, the same variables
should be measured in the pre-test and the post-test. Such a design
in not easy to implement, but is required if one wants to draw strong
conclusions about causality: if the treatment group, but not the other
two groups, improves on the measure(s) of interest, say ability in
mathematics, it is possible to conclude that the effect is specific to
some characteristic of the treatment group.
THE MAGIC BULLET? 103
Observers have not always been sympathetic to chess and some have
even noted a number of possible problems with its practice, which
can be categorised into two groups: non-psychiatric problems and
psychiatric problems. This chapter will address these two categories
in turn, with a focus on professional players.
NONPSYCHIATRIC PROBLEMS
A first obvious cost, which will actually be exacted for any activity
in which one wants to excel, is the huge investment of time that
is required. By necessity, this will leave little time for other activi-
ties. Anecdotal evidence indicates that many chess players aiming to
become professionals have left school early, often immediately after
obligatory school. And among the players who had a longer educa-
tion, many started studying at the university but never graduated.
This lack of completed education puts professional chess players at a
disadvantage if they decide, later in their life, to switch careers.
Although rich in excitement, competitive arousal and fun, a pro-
fessional chess player’s life is also very stressful. There is the obvious
tension caused by competitive games, which last for several hours,
and tournaments, which last for several days. In addition, few players
108 C O S T S O F P L A Y I N G C H E S S
PSYCHIATRIC PROBLEMS
A common assumption in newspapers, movies and the chess liter-
ature is that there is a close link between chess and madness. For
example, the 2011 documentary Bobby Fischer against the World played
with the themes of chess, madness and genius. In Chess: A Novel, Ste-
fan Zweig notes that is hard to imagine “a man of intelligence who,
without going mad, again and again, over ten, twenty, thirty, forty
years, applies the whole elastic power of his thinking to the ridiculous
goal of backing a wooden king into the corner of a wooden board!”
The issue of madness is also widespread in the chess psychology
literature, and was certainly a central theme in two books written by
psychoanalyst Reuben Fine, The Psychology of the Chess Player and Bobby
Fischer’s Conquest of the World’s Chess Championship. Understanding the pre-
sumed link between madness and genius is not unique to chess, and
has also been attempted with respect to creativity in music, literature
and the visual arts, in what is known as the “mad genius” hypothesis.
There has been much hype in this literature, but some of the cases
often discussed are rather straightforward, albeit tragic. For exam-
ple, there is much mention of an episode where Steinitz was briefly
admitted in a psychiatric hospital, amongst other things because he
thought he had invented a wireless phone. He was released rapidly
and able to conduct a reasonably normal life after, playing in chess
tournaments. A few weeks before his death, he was again hospital-
ised. One account reports that he wanted to play against God through
electric communication and give Him the advantage of a pawn and
first move. Fine hypothesises that Steinitz was suffering from organic
senile psychosis, with some delusional themes relating to his defeat
in his world championship match against Lasker.
Similarly, Alekhine was known to drink heavily, which led him
to extremes such as being drunk in some of the games of the world
championship against Euwe and urinating in public during a simul-
taneous display. He could also be extremely violent, once destroying
the furniture of his hotel room after losing a game. Whether this
denotes mental illness is debatable, however. This being said, it is
110 C O S T S O F P L A Y I N G C H E S S
likely the case that some top players suffered from psychiatric dis-
eases. The most often discussed in the literature are Paul Morphy and
Bobby Fischer.
Paul Morphy
Paul Charles Morphy was born in 1837. After having learnt to play
chess by watching games between his father and uncle, he was one
of the best players in New Orleans when he was just 9 years old. At
12, he beat Hungarian master Johann Löwenthal 3–0. However, he
did not play much in the following years, devoting himself to his
college and later law studies, which he completed brilliantly. When
he returned to chess at the age of 20, he won all his games in the first
US championship. Since he was not allowed to practice law given his
young age, he travelled to Europe and challenged the best players of
the world, defeating most of them comprehensively, including Adolf
Andersen, the de facto world champion.
After his return to Louisiana in 1859 at the age of 22, he totally
stopped playing chess. He intended to focus on his career as a lawyer,
but the American Civil War (1861–1865) thwarted this plan. He did
not manage to restart his law practice after the end of the war, partly
because his customers were more interested in discussing chess with
him than their legal matters. In spite of this, he lived a comfortable life
thanks to the considerable fortune he had inherited from his family.
He died at the age of 47 from a stroke, after having taken a cold bath
immediately after a long walk on a hot day.
His life was highly organised: daily walk at noon along New
Orleans’ Canal Street, afternoon with his mother and performance
at the opera in the evening. In addition to the fact that he never prac-
ticed as a lawyer after the war, some aspects of his life might denote
symptoms of psychosis. He incorrectly believed that his brother-in-
law had stolen his fortune and wanted to poison him. He methodi-
cally organised his shoes in a half-circle. Occasionally, he would stop
during his daily walk and rather rudely stare at beautiful women. He
also apparently had the habit of walking on his veranda declaiming, in
C O S T S O F P L AY I N G C H E S S 111
French, “He will plant the flag of Castille on the walls of Madrid with
the cry of the city won and the little king will go away all abashed”.
However, when his family considered placing him in a sanatorium,
Morphy so eloquently defended his legal rights and his sanity that
he was sent home.
According to Fine, these idiosyncrasies support the hypothesis that
Morphy suffered from paranoia. Following Ernest Jones’s earlier anal-
yses (see Chapter 5), he argues that Morphy’s symptoms were caused
by his rivalry with his father, which was sublimated by playing chess.
He notes that Morphy’s international successes started one year after
the early and unexpected death of his father. In addition, Fine takes
Morphy’s refusal to consider chess as a profession as an impossibil-
ity to accept reality and his extreme confidence in his chess abilities
as a sign of exhibitionism. He also proposes that, in his adolescent
years, chess succeeded in protected him from psychosis. When he
stopped playing chess, this defence disappeared, triggering a psy-
chotic regression.
The soundness of this analysis is debatable. Several biographical
details used by Jones and Fine are actually incorrect, as is clear from
reading David Lawson’s very detailed biography of Morphy. Alterna-
tive explanations are not considered, including: the effect of the Civil
War on Morphy’s mental health, the presence of hereditary causes to
his paranoia and the possibility that his decision to stop playing chess
might have been caused by the very bad press that chess had in the
US at the time, being considered as a type of gambling.
Bobby Fischer
These hypotheses about the link between chess and madness are
intriguing, but also controversial. A first problem is that the quality
of the data is poor – anecdotes, personal observations by colleagues,
press reports, etc. It is noticeable that Fine, although he personally
knew several of the champions he discussed in his books, never car-
ried out a formal and detailed clinical evaluation of any of them.
Second, there is no consideration of base rates. Noting the pres-
ence of paranoid personalities, for example, is one thing, but before
speculating on any link between genius and madness in chess, one
needs to establish that the prevalence of paranoid personalities (or
other psychiatric illnesses) in chess is higher than in the general
population. Given that the data span two centuries, this is obviously
difficult to do, as the prevalence is likely to have changed during this
period of time. Similarly, geographical and cultural differences should
be taken into account.
Third, and related to the previous point, it is obvious that some of
the world champions have lived in extremely difficult conditions –
think of the financial duress faced by Steinitz most of his life, and
Morphy’s experience of the Civil War. Psychoanalytical analyses do
not take this into account.
Finally, many of the analyses presented in this chapter are based
on psychoanalysis, which has been discredited scientifically and only
plays a minor role today in psychology and psychiatry. Specifically,
C O S T S O F P L AY I N G C H E S S 115
We have played a long game and have now reached its final stage. It is
time to draw together the different strands of this book.
Researchers now have a fairly good understanding of the topics
covered in the first four chapters. We know that perception is cen-
tral to expertise in chess, knowledge is mostly stored in long-term
memory as perceptual chunks linked to possible actions and search is
highly selective. We also know that skill in chess mostly finds it origin
in a combination of innate factors and dedicated practice. The topics
in the second half of the book have been less researched, and less is
known about them. They also tend to be more applied and harder to
study experimentally.
One theme that has been present in nearly all chapters is that of
bounded rationality. Developed in the 1950s by Herbert Simon –
whom we have met in several chapters of this book – bounded ratio-
nality proposes that humans make decisions that are good enough,
but not optimal, because of the limits imposed by their cognitive
resources. On the one hand, chess grandmasters choose remarkably
good moves: they are nearly always within the first three best moves
selected by the top computer engines, and often match the best one.
On the other hand, this book has presented substantial empirical
evidence that their rationality is limited: their search is extremely
118 E N D G A M E
Burgoyne, A. P., Sala, G., Gobet, F., Macnamara, B. N., Campitelli, G., & Hambrick,
D. Z. (2016). The relationship between cognitive ability and chess skill: A
comprehensive meta-analysis. Intelligence, 59, 72–83.
This meta-analysis clearly shows that expertise in chess is related to
intelligence.
Campitelli, G., & Gobet, F. (2004). Adaptive expert decision making: Skilled
chess players search more and deeper. Journal of the International Computer
Games Association, 27, 209–216.
In the experiment reported in this paper, participants had to solve fiend-
ishly complicated problems. The verbal protocols revealed very clear-cut
skill differences in the amount of search that was carried out.
Campitelli, G., Gobet, F., & Bilalić, M. (2014). Cognitive processes and devel-
opment of chess genius: An integrative approach. In D. K. Simonton
(Ed.), The Wiley handbook of genius (pp. 350–374). Chichester, UK: Wiley-
Blackwell.
This book chapter describes a model of the interaction of practice and
talent in chess and derives predictions through mathematical simulations.
Charness, N. (1976). Memory for chess positions: Resistance to interference.
Journal of Experimental Psychology: Human Learning and Memory, 2, 641–653.
This article presents several experiments showing that memory for briefly
presented chess positions is unaffected by interpolated tasks. This result
challenges chunking theory.
Charness, N., Reingold, E. M., Pomplun, M., & Stampe, D. M. (2001). The per-
ceptual aspect of skilled performance in chess: Evidence from eye move-
ments. Memory & Cognition, 29, 1146–1152.
This article provides interesting data on the role of eye movements when
chess players try to find the best move in a position.
Chassy, P. & Gobet, F. (2010). Speed of expertise acquisition depends upon
inherited factors. Talent Development and Excellence, 2, 17–27.
This article proposes a genetic hypothesis that accounts for inter-individ-
ual differences in the acquisition of expertise.
Dartigues, J. F., et al. (2013). Playing board games, cognitive decline and demen-
tia: A French population-based cohort study. BMJ Open, 3.
This is a large study looking at the possible benefits of playing board
games and other games for reducing the risk of dementia.
De Groot, A. D. (1978). Thought and choice in chess (2nd English ed.; 1st Dutch ed. in
1946). The Hague: Mouton Publishers.
F U RT H E R R E A D I N G 121
Gobet, F., & Campitelli, G. (2006). Education and chess: A critical review. In
T. Redman (Ed.), Chess and education: Selected essays from the Koltanowski conference (pp.
124–143). Dallas, TX: Chess Program at the University of Texas at Dallas.
This book chapter reviews the research using chess for improving cogni-
tive abilities and educational achievement. It provides a detailed discussion
on the ideal methodology to use for carrying out this type of research.
Gobet, F., & Campitelli, G. (2007). The role of domain-specific practice, handed-
ness and starting age in chess. Developmental Psychology, 43, 159–172.
This is one of the first papers to challenge the deliberate practice frame-
work with the kind of empirical data used by this approach. Its contribu-
tion was also to combine data on deliberate practice with data on talent.
Gobet, F., & Chassy, P. (2009). Expertise and intuition: A tale of three theories.
Minds & Machines, 19, 151–180.
This article presents a critique of Simon and Chase’s and Dreyfus and
Dreyfus’s theories of intuition, and proposes a new theory, based on tem-
plate theory (see Chapter 2).
Gobet, F., de Voogt, A. J., & Retschitzki, J. (2004). Moves in mind. Hove, UK: Psy-
chology Press.
This book presents a systematic review of the psychology of board games,
with particular attention to chess.
Gobet, F., & Jansen, P. J. (2006). Training in chess: A scientific approach. In
T. Redman (Ed.), Chess and education: Selected essays from the Koltanowski conference
(pp. 81–97). Dallas, TX: Chess Program at the University of Texas at Dallas.
This book chapter uses results in cognitive psychology to develop prin-
ciples of chess instruction.
Gobet, F., Lane, P. C. R., Croker, S., Cheng, P. C. H., Jones, G., Oliver, I., & Pine,
J. M. (2001). Chunking mechanisms in human learning. Trends in Cognitive
Sciences, 5, 236–243.
This is a review showing the general role of chunking mechanisms in
human cognition, beyond chess. It provides a brief presentation of the
CHREST model.
Gobet, F., & Simon, H. A. (1996). The roles of recognition processes and look-
ahead search in time-constrained expert problem solving: Evidence from
grandmaster level chess. Psychological Science, 7, 52–55.
This paper analyses world champion Garry Kasparov’s simultaneous
exhibitions against national teams and the team of Hamburg. It takes
F U RT H E R R E A D I N G 123
Maass, A., D’Ettole, C., & Cadinu, M. (2008). Checkmate? The role of gender
stereotypes in the ultimate intellectual sport. European Journal of Social Psychol-
ogy, 38, 231–245.
The article argues that male domination in chess is due to gender
stereotypes.
Moul, C. C., & Nye, J. V. (2006). Did the Soviets collude? A statistical analy-
sis of championship chess 1940–64. Retrieved from http://dx.doi.
org/10.2139/ssrn.905612.
An interesting statistical analysis that supports Bobby Fischer’s claim that
Soviet grandmasters essentially cheated against Western players.
Murray, H. J. R. (1952). A history of chess (reprint of the 1913 ed.). New York, NY:
Skyhorse Publishing Press.
First published in 1913, this is the definitive reference on the history of
chess.
Polgár, L. (2017). Raise a genius! (Original ed. in Hungarian, 1989). Translation
and copyright: Gordon Tisher.
In this book, László Pogar presents his views on education and describes
the “experiment” he and his wife carried out on their three daughters.
For many years, the book was available only in Hungarian and Esperanto.
It can be downloaded at http://slatestarcodex.com/Stuff/genius.pdf.
Saariluoma, P. (1992). Error in chess: The apperception-restructuring view. Psy-
chological Research, 54, 17–26.
This paper describes elegant studies where errors are induced
experimentally.
Saariluoma, P. (1995). Chess players’ thinking: A cognitive psychological approach. London:
Routledge.
This book discusses the psychology of chess, with a focus on the author’s
research. Experiments on blindfold chess are discussed in some detail.
Sala, G., & Gobet, F. (2016). Do the benefits of chess instruction transfer to
academic and cognitive skills? A meta-analysis. Educational Research Review,
46–57.
This meta-analysis statistically combines the results of 24 studies.
Silver, D., et al. (2016). Mastering the game of Go with deep neural networks
and tree search. Nature, 529, 484–489.
AlphaGo, which is described in this paper, revolutionised the way com-
puters play Go.
F U RT H E R R E A D I N G 125
Simon, H. A., & Chase, W. G. (1973). Skill in chess. American Scientist, 61, 393–403.
In this classic paper, Simon and Chase provide an overview of the seminal
research they carried out on chess.
Stafford, T. (2018). Female chess players outperform expectations when playing
men. Psychological Science, 29, 429–436.
Using data from a large database, this article argues that women do not
suffer from any stereotype threat, but rather play better than expected
when facing men.